Posted at the Institutional Resources for Unique Collection and Academic Archives at Tokyo Dental College, Available from http://ir.tdc.ac.jp/
Title
Expression of BMI1 and ZEB1 in epithelial‑
mesenchymal transition of tongue squamous cell carcinoma
Author(s) 栗原, 絹枝 Journal , (): ‑
URL http://hdl.handle.net/10130/3419 Right
1
Expression of BMI1 and ZEB1 in epithelial-mesenchymal transition of tongue squamous cell carcinoma
Kinue Kurihara
Department of Oral Medicine, Oral and Maxillofacial Surgery,
Tokyo Dental College
Chief Director
Akira Katakura
2
Abstract
Introduction: The Epithelial-mesenchymal transition (EMT), the crucial event for
the invasion and progression of epithelial carcinogenesis, induces stem-like
properties for epithelial cells. Expression of BMI1 which controls self-renewal in
stem cells and ZEB1, a transcription factor which regulates EMT were studied on
the role in the carcinogenesis process of tongue squamous cell carcinoma
(TSCC).
Material and methods: Collagen invasion assay using 2 TSCC cell lines and 64
tongue tissue specimens (32 carcinomas and 32 dysplasias) were used the
materials of this study. We assessed protein and mRNA expression levels of
BMI1, ZEB1, Vimentin and E-cadherin in both cell lines and tumor tissues.
Results: Both of the protein and mRNA expression of BMI1 and ZEB1 occurred
at the invasion of TSCC. The elevated levels of BMI1 and ZEB1 were
accompanied by downregulation of E-cadherin and upregulation of Vimentin at
the invasive front, induced of EMT both in vivo and in vitro.
Conclusions: We demonstrate that BMI1 and ZEB1 are important factors in
associated with promotion of EMT and invasion of TSCC.
3
Introduction
Head and neck squamous cell carcinoma (HNSCC) including oral cancer is
the sixth most prevalent cancer worldwide and accounts for approximately
8-10% of all cancers in Southeast Asia [1, 2]. At present, despite improvements
in the diagnosis and management of HNSCC, long-term survival rates have
improved only marginally over the past decade [3]. To improve the survival rate
of HNSCC patients, investigations the underlying molecular and phenotypic
events associated with head and neck squamous tumorigenesis to identify
biomarkers for early detection and prognostic stratifications are needed.
Recently studies have suggested that the persistent survival of cancer stem cells
(CSCs), also known as tumor-initiating cells, may contribute to the aggression
and recurrence of HNSCC [4, 5, 6, 7]. These CSCs are key contributors to
radioresistance and chemoresistance and are responsible for tumor progression
and recurrence after conventional therapy (radiotherapy and chemotherapy) [4,
8].
The epithelial-mesenchymal transition (EMT) is a key developmental program
4
that CSCs is often activated during cancer development [9, 10]. The occurrence
of EMT in cells may lead to the number of changes including loss of cancer cell
polarity and downregulation of epithelial cell markers, loss of cell-cell connection,
gain of mesenchymal phenotypes along with genetic/epigenetic modifications of
various genes. Published reports suggest a direct link between the EMT and the
gain of CSCs-like properties [11]. This process is thought to be a critical step in
the induction of tumor metastasis and malignancy [12].
BMI1 (B-lymphoma Moloney murine leukemia virus insertion region-1), a
member of polycomb group (PcG) genes, is considered to be pivotal in
regulating a stemness-related gene maintaining the self-renewal capacity of
stem cells through promoting chromatin modifications, and is also known to be
deregulated in various human cancers [13, 14, 15, 16]. BMI1 is e a prognostic
marker in prostate cancer [17], breast cancer [18], ovarian cancer [19], cervical
cancer [20], colorectal cancer [21], lung cancer [22], esophageal cancer [23],
gastric cancer [24], and nasopharyngeal cancer [13]. However, the role of BMI1
in maintaining self-renewal properties and tumorigenicity in HNSCC or
HNSCC-derived cancer stem cells (CSCs) has yet to be clarified.
5
ZEB1, a member of the zinc-finger transcription factor family, is one of the
master regulators of EMT that mediates invasiveness as well as metastasis in
many different types of malignant tumors. ZEB1 induces EMT by suppressing
the expression of E-cadherin and contribute to the progression of malignant
cancer [25]. ZEB1 is a good predictor of prognosis in breast cancer [26], lung
cancer [27], colorectal cancer [28], and esophageal cancer [29]. Extensive
studies have revealed that several transcription factors such as ZEB1 work
together to regulate the EMT program [30]. However, the role of ZEB1 in
HNSCC remains unclear.
Here, we studied several biomarkers, BMI1 and ZEB1, Vimentin, E-cadherin
associated with EMT in tongue squamous cell carcinoma (TSCC) cells and
tumor specimens to determine their relation to the invasion and progression of
TSCC which accounts for approximately 60% of oral squamous cell carcinoma
and clarified the significance of BMI1 and ZEB1 in TSCC.
6
Materials and methods
Carcinoma cell lines and isolation of fibroblasts
Two human tongue squamous cell carcinoma cell lines, TOSCa-2S,
TOSCa-23, and human fibroblasts [31]. Human fibroblasts were collected from
human oral specimens and used as stromal cells for this assay.
Collagen gel invasion assay and immunohistochemistry of TSCC cells
To conduct the collagen gel invasion assay, we used a 3-dimensional collagen
gel culture. Insert chambers with 8 μm pore filter were placed in six 35 mm
culture dishes.
First, a collagen solution was poured into insert chambers and incubated at 37℃
for 30 minutes to solidify the gel. Secondly, eight volumes of acid-soluble 0.3%
type Ⅰ collagen solution (Cellmatrix typeⅠ -A, pH 3), one volume of 10×
concentrated minimum essential medium, and 1 volume of reconstruction buffer
(2.2g of sodium bicarbonate and 4.77g of HEPES dissolved in 100mL of 0.05N
sodium hydroxide) were mixed. Fibroblasts were added to this solution at a
7
density of 1×105/ml, then 2ml of this mixture containing fibroblasts was added to
the chamber on top of the solid collagen-only layer. After gelation, TSCC cells
suspension at a concentration of 1×106 cells per dish was spread on the gel.
After incubating for four weeks, both of TSCC cell lines were observed migrating
into the underlying gel. The whole collagen gel was fixed with 10% formalin
embedded in paraffin, 4µm thinsectioned vertically, stained with
hematoxylin-eosin. For immunostaining, antigens were retrieved by heating at
120℃ for 20 minutes, cancer cells were identified with Vimentin, E-cadherin,
BMI1 and ZEB1 antibodies (Fig 1).
Analysis of TSCC cell invasion
The linear borderline formed at the contact points between the cells and the
gel was corresponding to a basement membrane-like structure. We judged
TSCC cells organized a stratified layer on the gel or contacting with the
basement membrane as a preinvasion, and the downgrowth into the gel
separating from the basement membrane as an invasion. Four specimens for
each cell line were observed, preinvasive cells and invasive cell were counted
8
and examined the percentage of positive cells and correlation between the
group of preinvasive and invasive cells.
Tissue samples and patients
Tongue tissue specimens accessed at Oral Pathology of Showa University
from 1997 to 2011 were used the materials of this study. A total of 47 patients
were eligible of inclusion (24 men and 23 women, median age 58 years, range
30-83 years). All patients had undergone resection of the tongue primary tumor
and this study did not include any patient with a distant metastasis and any who
had received preoperative therapy.
This study was approved by the Committee on Ethics, Oral Pathology of Showa
University, adhered to the principles of the Declaration of Helsinki, and all of the
samples were obtained after the patients had provided their informed consent.
(Permit number 8, November 2, 2001)
Tongue tissues were surgically resected from patients and the hematoxylin and
eosin-stained slides were reviewed. The tissues were immediately sent to
Pathology lab, stored at -80℃ for immunohistochemistry and real-time RT-PCR.
9
Sixty-four lesions obtained from fresh-frozen tongue tissue specimens were
consisted 32 primary invasive tongue cancer ( 15 early invasive cancer which do
not invade the muscle layer and 17 advanced invasive cancer which invade the
muscle layer) and 32 dysplasia ( 14 mild dysplasia and 18 moderate-severe
dysplasia). There were some specimens with invasive cancer and dysplasia
simultaneously. The original histological sections and immunostainings were
evaluated by a single pathologist without knowledge of clinical data.
Immunohistochemistry for Tongue tissues
The frozen tissue slides were cut into 4μm-thick sections, fixed in 4%
paraformaldehyde and treated with 3% hydrogen peroxide in methanol for 10
minutes to block the endogenous peroxidase. Immunostaining was performed
with a mouse monoclonal E-cadherin antibody (Santa Cruz Biotechnology,
SC-8426, diluted 1:100), or rabbit polyclonal BMI1 antibody (Cosmo Bio
Epitomics, T3421, 1:100), or rabbit polyclonal ZEB1 antibody (Santa Cruz
Biotechnology, SC-25388, 1:100) overnight at 4℃, or mouse monoclonal
Vimentin antibody (Dako, #M0725; 1:200) for 30 minutes at room temperature.
10
After rinsing in phosphate buffered saline, sections were incubated with
biotinylated secondary antibody. Detection was performed with
diaminobenzidine (DAB) and counterstained with Mayer hematoxylin.
Assessment of immunohistochemical staining
Immunostainings were observed in the parabasal and basal cell layer of the
normal squamous epithelium and dysplasia, the outermost layer of cancer nest
in the invasive front of the invaded cancer. The semi-quantitative analysis of the
stained sections was done by light–microscopy according to the immunoreactive
scoring (IRS) system by Remmele and Stegner. Sections were examined at
400×magnification and the staining intensity (SI) was evaluated by comparison
with adjacent normal epithelia, which served as a reference for moderate
intensity (M). Tumor staining less intense than the basal layer of adjacent normal
epithelia was categorized as weak intensity (W), more intense staining was
categorized as strong intensity (S), and no staining was categorized as negative
(N). We also calculated the percentages of cells with different staining intensities
to have ten spots for each slide and the predominant intensity was recorded for
11
each tumor. Based on the percentage of positive cells (PP), stainings were
classified into five grades: grade 0(0%), grade 1 (0-10%), grade 2 (11-50%),
grade 3 (51-80%) and grade 4 (80-100%). The product of SI and PP was the IRS.
The IRS with points from 0 to 12 was adapted to an additional 3-points IRS
classification (Table 1).
Laser microdissection
The 8μm-thick sliced sample was fixed in 95% ethanol for 5min, and then
washed with 70% ethanol, and stained with LCM frozen section staining kit
(AMBION). We procured a few hundred cells from cancer tissue and epithelial
dysplasia and adjacent normal epithelia in each 15 cases using laser
microdissection （PALM MicroBeam, ZEISS）for analysis of gene expression.
The mRNA expression were collected from the parabasal and basal cell layer of
the normal squamous epithelium and dysplasia, cancer nest in the invasive front
of the invaded cancer. The microdissected cells within the cap were covered
with 100 μl buffer solution and the capped tube was vortexed. Total mRNA was
independently extracted from each population of laser-microdissected cells.
12
Quantitative real-time reverse transcription –PCR
Total mRNA was extracted with RNeasy Plus Micro Kit (QIAGEN) from frozen,
microdissected samples of cancer tissue, epithelial dysplasia and adjacent
normal epithelia, according to the manufacturer’s instructions. mRNA was
reverse transcribed with SuperScript VILO Master Mix (Invitrogen) and cDNA
synthesis was performed. Quantitative real-time PCR was performed with an
ABI PRISM 7500 Fast Real-Time PCR System (Applied Biosystems).
The amplification profile was denatured at 95℃ for 10 min, followed by 50
cycles of denaturation at 95℃ for 15 sec, annealing at 60℃ for 1 min. The
expression levels were quantified using the following Vimentin primer
(Hs00185584_m1, Taqman, Applied Biosystems), E-cadherin primer
(Hs01023894_m1, Taqman, Applied Biosystems), BMI1 primer
(Hs00180411_m1, Taqman, Applied Biosystems), and ZEB1 primer
(Hs00232783_m1, Taqman, Applied Biosystems). The geometric mean of the
GAPDH (glyceraldehydes-3-phosphate dehydrogenase housekeeping gene)
was used as an internal control to normalize the variability in expression levels.
The comparative cycle threshold (CT) method was applied to quantify the
13
expression levels of mRNAs. The relative amount of each marker was calculated
using the equation 2-ΔC T where ΔCT = (CTX– CTGAPDH)
Statistical analysis
Comparison the protein expression levels according to the IRS between two
groups was used for the Kruskal Wallis test and the Mann-Whitney post hoc,
and comparison the mRNA expressions was used for the non-repeated
measures ANOVA and post hoc. The correlation between the expressions of
several biomarkers and TSCC cells was evaluated with the Chi-square test and
the Fisher’s exact test, spearman’s correlation. All the statistical analyses were
performed using modified EZR (The R Foundation for Statistical Computing,
Perugia, Italy) software programs. Two-tailed P-values were calculated, and
P<0.05, <0.01 was considered to indicate a statistically significant result.
14
Results
Protein expression of TSCC cells in invasion assay
In collagen gel invasion assay, we evaluated protein expressions in early
invasion carcinogenesis of TSCC cell (TOSCa-2S, TOSCa-23). Only 2.5% of the
preinvasive cells, as many as 70.0% of the invasive cells were defined as
manifesting high Vimentin expression. We distinguished between fibroblasts and
invasive cancer cells in collagen gel stained with Vimentin. 72.3% of the
preinvasive cells, only 3.6% of the invasive cells were defined as manifesting
high E-cadherin expression. 62.4% of the preinvasive cells, 73.0% of the
invasive cells were defined as manifesting high BMI1 expression, and 60.2% of
the preinvasive cells, 73.5% of the invasive cells were defined as manifesting
high ZEB1 expression. Protein expressions of all markers were significantly
difference between preinvasive cells and invasive cells (P<0.001, P<0.01; Fig 2).
Protein and mRNA expression of tongue cancer and dysplasia specimens
We compared of immunohistochemical expression among the five groups,
15
normal squamous epithelium, mild dysplasia, moderate-severe dysplasia, early
invasive cancer and advanced invasive cancer. BMI1 immunoexpression was
mainly localized to the nuclei of normal squamous epithelium and dysplasia but
detected in both the nuclei and cytoplasm of the majority of tongue cancer. BMI1
high immunoexpression (IRS-classification: 12) was observed in 84.4% (27 of
32) of invasive cancer. ZEB1 immunoexpression was detected in cytoplasm of
all samples and ZEB1 high immunoexpression was observed in 50.0% (16 of
32) of invasive cancer. (Fig 3)
E-cadherin protein expression level was significantly decreased in
moderate-severe dysplasia and invasive cancer compared with adjacent normal
squamous epithelia (P=0.0021, P<0.001, P<0.0001; Fig 4). Vimentin and BMI1
protein expression levels were significantly increased in invasive cancer,
including early and advanced invasive cancer, (P=0.0074, <0.001, P=0.04,
0.0001; Fig 4), and ZEB1 protein expression levels was significantly increased
only in advanced invasive cancer (P=0.014; Fig 4), compared with adjacent
normal squamous epithelia.
E-cadherin mRNA expression level was significantly decreased in
16
moderate-severe dysplasia and invasive cancer compared with adjacent normal
squamous epithelia (P=0.039, P=0.019; Fig 5). Vimentin, BMI1 and ZEB1 mRNA
expression levels were significantly increased in invasive cancer compared with
adjacent normal squamous epithelia (P=0.047, P=0.036, P=0.045; Fig 5).
Elevated levels of BMI1 were accompanied by downregulation of E-cadherin and
upregulation of Vimentin at the invasive front, demonstrated a significant
negative correlation between BMI1 and E-cadherin protein and mRNA
expressions (P= 0.0097; Fig 6A, P= 0.018; Fig 6B), and a significant positive
correlation between BMI1 and Vimentin protein and mRNA expressions.
(P=0.035; Fig 6A, P=0.0008; Fig 6B). There was a significant positive correlation
between ZEB1 and Vimentin mRNA expressions, between BMI1 and ZEB1
mRNA expressions (P<0.001, P=0.024; Fig 6B).
Discussion
EMTs are encountered in three distinct biological settings [32]. Type 1 of EMTs
are associated with implantation, embryonic gastrulation, gives rise to the
17
mesoderm and endoderm and to mobile neural crest cells and organ
development are organized. Type 2 of EMTs are re-engaged in the context of
inflammation and fibrosis. These EMTs continued to occur until infections are
removed over extended periods of time and the tissue is repaired. Type 3 of
EMTs occur in the context of tumor growth and cancer progression, and the
epithelia can transform into cancer cells that later undergo EMTs that enable
invasion and metastasis. All EMTs have in common and E-cadherin
transcriptional repression is characteristic of EMT. Given that the downregulation
of E-cadherin is tightly associated with EMT program and the invasion of cancer,
it is worth remembering that E-cadherin loss has long been associated with the
progression of papilloma to invasive carcinoma [33].
In the present study, we examined the possible involvement in tongue
carcinogenesis by comparison of BMI1 and ZEB1, E-cadherin, Vimentin protein
and mRNA expression levels in normal, dysplasia and TSCC tissues and TSCC
cells.
In TSCC cells invasion assay, we demonstrated that downregulation E-cadherin
protein expression observed in 96.4% of the invasive TSCC cells were
18
demonstrated the promotion of EMT program and TSCC cells invasion. We
revealed that BMI1 and ZEB1 protein overexpression were observed in 73% of
the invasive TSCC cells compared with 60% of the preinvasive TSCC cells. This
finding indicated that BMI1 and ZEB1 protein overexpression are associated
with the EMT program and the TSCC cells invasion.
In TSCC tissues, BMI1 protein and mRNA expression levels were significantly
increased in invasive tongue cancer, including early and advanced invasive
cancer, compared with adjacent normal squamous epithelia. At the invasive front,
elevated levels of BMI1 were accompanied by downregulation of E-cadherin and
upregulation of Vimentin, demonstrated a significant negative correlation
between BMI1 and E-cadherin expressions, a significant positive correlation
between BMI1 and Vimentin expressions. These data suggest that BMI1
overexpression at both mRNA and protein levels was involved in the invasion
and progression of TSCC. These findings are in agreement with previous
studies of other epithelial malignancies and further support an important role for
BMI1 activation in the downregulation of E-cadherin and the induction of EMT.
For instance, Song et al. demonstrated that Bmi-1 mRNA and protein
19
expressions were found to correlate with the invasion of nasopharyngeal
carcinoma [13]. Yang et al. showed that BMI1 is essential for EMT during tumor
development in head and neck cancer patients [14]. MK Kang et al. showed
Bmi-1 overexpression was observed in 100% of the preneoplastic oral mucosal
tissue which included those with mild, moderate or severe epithelial dysplasia
[34]. This divergence from our results may be due to the different
pathophysiology of oral squamous cell carcinoma such as non-keratinising type
or ketatinising of carcinoma, but our material is a keratinizing SCC of tongue
carcinoma. The former study constantly presented in a small series of patients
(N=8, 10) with oral dysplastic and carcinoma tissue, but our research data
consist of 64 cases. Interestingly, V Hayry et al. showed a significant negative
correlation between Bmi-1 protein expression and the recurrence of tongue
cancer [35]. This divergence from our results may be due to selecting part of
tissue which detached with 1mm punch from the surface epithelium and a
central of the tumor and the invading front. Balasubramanian et al. reported the
expression of BMI1 in basal and suprabasal keratinocytes and is not present in
surface epithelium [36]. Recently study showed that in the invading front, BMI is
20
highly enriched in CSCs but there is not Bmi-1 expression in all cancer cells [37].
In the present study, we just evaluate BMI1 protein expression and mRNA
expression of cancer nest cells in the invasive front.
The overexpression of ZEB1 was observed in colorectal cancer and esophageal
cancer [28, 29], suggesting an important role in tumorigenesis. EMT-induced
ZEB1 was previously reported to be associated with cancer progression [38]. In
the present study, we confirmed that ZEB1 protein and mRNA expression levels
were significantly increased in advanced invasive cancer compared with
adjacent normal squamous epithelia, consistent with the results of previous
studies and suggested to be associated with cancer progression.
On the other hand, ZEB1 was found to be responsible for downregulation of
basal membrane constituents at the invasive front of colorectal carcinoma [29]
and Takehiko Y et al. showed that in esophageal SCC, ZEB1 targeting by
miR-150 could suppress E-cadherin repression, Vimentin expression, migration
ability, and tumorigenicity [30]. In this study, we indicated that there was a
significant positive correlation between ZEB1 and Vimentin mRNA expressions,
but was a no significant correlation between ZEB1 and E-cadherin expressions.
21
A hallmark for EMT is the loss of E-cadherin expression, however, ZEB1 with the
EMT process of TSCC may not suppress E-cadherin expression.
In this study, we demonstrated that elevated level of BMI1 was accompanied by
downregulation of E-cadherin and upregulation of Vimentin, and elevated level
of ZEB1 was accompanied upregulation of Vimentin at the invasive front of
TSCC.
The recent study found that the EMT has important roles in cancer invasion and
metastasis and cancer stem cell properties [4, 9]. The study demonstrated that
ZEB1 and ZEB2 are key modulators of CSC properties in HNC, including EMT,
metastasis, and drug resistance [39]. At this time, this study indicated that
activation of BMI1, stem cell-like marker, is associated with promotion of EMT
and invasion in TSCC and further investigation into the role and mechanisms of
BMI1 and ZEB1 in TSCC is require.
In conclusion, BMI1 and ZEB1 are important factors in associated with
promotion of EMT and invasion of TSCC.
22 Acknowledgements
We appreciate the technical help from the Showa University Pathology
Department.
References
1. Haddad RI, Shin DM. Recent advances in head and neck cancer. N Engl J Med. 2008;359:1143-54.
2. Chen YJ, Chang JT, Liao CT, Wang HM, Yen TC, Chiu CC et al. Head and neck cancer in the betel quid chewing area: recent advances in molecular carcinogenesis. Cancer Sci. 2008;99:1507-14
3. Gibson MK, Forastiere AA. Reassessment of the role of induction chemotherapy for head and neck cancer. Lancet Oncol. 2006;7:565-74.
4. Zhang Z, Filho MS, Nör JE. The biology of head and neck cancer stem cells.
Oral Oncol. 2012;48:1-9.
5. Rosen JM, Jordan CT. The increasing complexity of the cancer stem cell paradigm. Science. 2009;324:1670-3.
6. Lo WL, Yu CC, Chiou GY, Chen YW, Huang PI, Chien CS et al.
MicroRNA-200c attenuates tumour growth and metastasis of presumptive head and neck squamous cell carcinoma stem cells. J Pathol.
2011;223:482-95.
7. Chen C, Zimmermann M, Tinhofer I, Kaufmann AM, Albers AE.
23
Epithelial-to-mesenchymal transition and cancer stem(-like) cells in head and neck squamous cell carcinoma Cancer Letters. 2012;338:47-56
8. Major AG, Pitty LP, Farah CS. Cancer stem cell markers in head and neck squamous cell carcinoma. Stem Cells Int. 2013;2013:319489:p1-13.
9. Thiery JP, Acloque H, Huang RY, Nieto MA.Epithelial-mesenchymal transitions in development and disease. Cell 2009;139:871-90
10. Raimondi C, Gianni W, Cortesi E, Gazzaniga P. Cancer stem cells and epithelial-mensenchymal transition: Revisiting minimal residual diseal. Curr Cancer Drug Targets 2010;10:496-508.
11. Polyak K, Weinberg RA. Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer.
2009;9:265-73.
12. Thiery JP, Sleeman JP. Complex networks orchestrate epithelial-mesenchymal transitions. Nat Rev Mol Cell Biol. 2006;7:131-42.
13. Song LB, Li J, Liao WT, Feng Y, Yu CP, Hu LJ et al. The polycomb group protein BMI1 represses the tumor suppressor PTEN and induces epithelial-mesenchymal transition in human nasopharyngeal epithelial cells.
J Clin Invest 2009;119:3626-36.
14. Yang MH, Hsu DS, Wang HW, Wang HJ, Lan HY, Yang WH et al. Bmi1 is essential in Twist1-induced epithelial-mesenchymal transition. Nat Cell Biol 2010;12:982-92.
15. Gil J, Bernard D, Peters G. Role of polycomb group proteins in stem cell
24
self-renewal and cancer. DNA Cell Biol. 2005;24:117-25.
16. Pasini D, Bracken AP, Helin K. Polycomb group proteins in cell cycle progression and cancer. Cell Cycle. 2004;3:396-400.
17. Crea F, Duhagon Serrat MA, Hurt EM, Thomas SB, Danesi R, Farrar WL.
BMI1 silencing enhances docetaxel activity and impairs antioxidant response in prostate cancer. Int J Cancer. 2011;128:1946-54.
18. Guo BH, Feng Y, Zhang R, Xu LH, Li MZ, Kung HF et al. Bmi-1 promotes invasion and metastasis, and its elevated expression is correlated with an advanced stage of breast cancer. Mol Cancer. 2011;10:10:p1-12
19. Wang E, Bhattacharyya S, Szabolcs A, Rodriguez-Aguayo C, Jennings NB, Lopez-Berestein G et al. Enhancing chemotherapy response with Bmi-1
silencing in ovarian cancer. PLoS One. 2011;6:e17918:p1-9.
20. Tong YQ, Liu B, Zheng HY, He YJ, Gu J, Li F et al. BMI-1 autoantibody as a new potential biomarker for cervical carcinoma. PLoS One.
2011;6:e27804:p1-7.
21. Kim JH, Yoon SY, Kim CN, Joo JH, Moon SK, Choe IS et al. The Bmi-1 oncoprotein is overexpressed in human colorectal cancer and correlates with the reduced p16INK4a/p14ARF proteins. Cancer Lett.
2004;203:217-24.
22. Vonlanthen S, Heighway J, Altermatt HJ, Gugger M, Kappeler A, Borner MM et al. The bmi-1 oncoprotein is differentially expressed in non-small cell lung cancer and correlates with INK4A-ARF locus expression. Br J Cancer.
25 2001;84:1372-6.
23. He XT, Cao XF, Ji L, Zhu B, Lv J, Wang DD et al. Association between Bmi1 and clinicopathological status of esophageal squamous cell carcinoma.
World J Gastroenterol. 2009;15:2389-94.
24. Liu JH, Song LB, Zhang X, Guo BH, Feng Y, Li XX et al. Bmi-1 expression predicts prognosis for patients with gastric carcinoma. J Surg Oncol.
2008;97:267-72.
25. Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest. 2009;119:1420-8
26. Soini Y, Tuhkanen H, Sironen R, Virtanen I, Kataja V, Auvinen P et al.
Transcription factors zeb1, twist and snai1 in breast carcinoma. BMC Cancer.
2011;11:11:p1-12
27. Gemmill RM, Roche J, Potiron VA, Nasarre P, Mitas M, Coldren CD et al.
ZEB1-responsive genes in non-small cell lung cancer. Cancer Lett.
2011;300:66-78.
28. Zhang GJ, Zhou T, Tian HP, Liu ZL, Xia SS. High expression of ZEB1 correlates with liver metastasis and poor prognosis in colorectal cancer.
Oncol Lett. 2013;5:564-8.
29. Yokobori T, Suzuki S, Tanaka N, Inose T, Sohda M, Sano A et al. MiR-150 is associated with poor prognosis in esophageal squamous cell carcinoma via targeting the EMT inducer ZEB1. Cancer Sci. 2013;104:48-54.
30. Gheldof A, Hulpiau P, van Roy F, De Craene B, Berx G. Evolutionary
26
functional analysis and molecular regulation of the ZEB transcription factors.
Cell Mol Life Sci. 2012;69:2527-41.
31. Nakamura E, Kozaki K, Tsuda H, Suzuki E, Pimkhaokham A, Yamamoto G et al. Frequent silencing of a putative tumor suppressor gene melatonin receptor 1 A (MTNR1A) in oral squamous-cell carcinoma. Cancer Sci. 2008 99:1390-400
32. Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G. A causal role for E-cadherin in the transition from adenoma to carcinoma. Nature 1998; 392:
190-3.
33. Nieto MA. The ins and outs of the epithelial to mesenchymal trasition in health and disease. Annu Rev Cell Dev Biol. 2011;27:347-76.
34. Kang MK, Kim RH, Kim SJ, Yip FK, Shin KH, Dimri GP et al. Elevated Bmi-1 expression is associated with dysplastic cell transformation during oral carcinogenesis and is required for cancer cell replication and survival. Br J Cancer. 2007;96:126-33.
35. Häyry V, Mäkinen LK, Atula T, Sariola H, Mäkitie A, Leivo I et al. Bmi-1 expression predicts prognosis in squamous cell carcinoma of the tongue. Br J Cancer. 2010;102:892-7
36. Balasubramanian S, Lee K, Adhikary G, Gopalakrishnan R, Rorke EA, Eckert RL. The Bmi-1 polycomb group gene in skin cancer: regulation of function by (-)-epigallocatechin-3-gallate. Nutr Rev. 2008;66:S65-8.
37. Siddique HR, Saleem M. Role of BMI1, a stem cell factor, in cancer
27
recurrence and chemoresistance: preclinical and clinical evidences. Stem Cells. 2012;30:372-8.
38. Chen ML, Liang LS, Wang XK. miR-200c inhibits invasion and migration in human colon cancer cells SW480/620 by targeting ZEB1. Clin Exp Metastasis. 2012;29:457-69.
39. Chu PY, Hu FW, Yu CC, Tsai LL, Yu CH, Wu BC et al.
Epithelial-mesenchymal transition transcription factor ZEB1/ZEB2 co-expression predicts poor prognosis and maintains tumor-initiating properties in head and neck cancer. Oral Oncol. 2013;49:34-41.
Table 1 Immunoreactive score(IRS) and IRS classification scoring systems Percentage of positive cells X intensity of staining = IRS (0-12)
0 = no positive cell 0 = no color reaction 1 = <10% positive cells 1 = mild reaction
2 = 10-50% positive cells 2 = moderate reaction 3 = 51-80% positive cells 3 = strong reaction 4 = >80% positive cells
IRS-points IRS-classification
0 - 4 0 less than normal negative 5 - 8 1 less than normal negative 9 - 12 2 more than normal positive
(Remmele and Stegner 1987)
H&E Vimentin E-cadheirn BMI1 ZEB1
×400
Fig 1 Immunoexpression of TSCC cells in collagen gel invasion assay
Vimentin expression
p-value
BMI1 expression
positive negative positive negative p-value
preinvasive cells 11 (2.5%) 426 (97.5%)
***< 0.001 preinvasive cells 305 (62.4%) 184 (37.6%)
**< 0.01
invasive cells 79 (70.0%) 34 (30.0%) invasive cells 135 (73.0%) 50 (27.0%)
E-cadherin expression
p-value
ZEB1 expression
positive negative positive negative p-value
preinvasive cells 334 (72.3%) 128 (27.7%)
***< 0.001 preinvasive cells 240 (60.2%) 159 (39.8%)
**< 0.01
invasive cells 4 (3.6%) 107 (96.4%) invasive cells 83 (73.5%) 30 (26.5%)
Fig 2 Protein expression of TSCC cells in collagen gel invasion assay
Fig 3 Immunoexpression of tongue tissues
N orm al epit helium
H&E Vimentin E-cadherin BMI1 ZEB1
Sev ere dy splas ia
×400
Adv anc ed inv as iv e carc inom a
Vimentin (IRS score)
Fig 4 Protein expressions of tongue tissues according to the IRS
N=14 N=18 N=15 N=17
Vimentin (IRS score)
* Dys: dysplasia, Ca: invasive carcinoma
N=4 N=4 N=4
N=10
Fig 5 mRNA expressions of tongue tissues
* Dys: dysplasia, Ca: invasive carcinoma
Fig 6 Correlation of the IHC grading of protein expressions (A) and Correlation of mRNA expressions (B)
P=0.018 P=0.0008
P=0.024 P<0.001
A B
Bmi-1 expression *P-value
negative positive
Grade 0 1 2
E-cadherin expression
negative 3 11 36 0.0097
positive 3 8 6
Bmi-1 expression *P-value
negative positive
Grade 0 1 2
Vimentin expression
negative 5 14 22 0.035
positive 1 5 20
